MaidSafe Dev Update :safe: 21st April 2016

Not sure it should fill and die. These tests are to gather the churn info so depends hoe many join how fast and upload what. For now we just want to check our own logs to validate some churn figures. So we cannot tell what’s happening only what happens looking at logs in the morning.

3 Likes

These are log entries, the errors are developer related info where we check for various events. They are not user level errors (i.e. not mistakes) but debug information.

3 Likes

OK, there appear to be two good lines in the middle, as noted.

“tcp_acceptor_port”: 61234,

thats a valid line for example. set your numbers accordingly to what you need.

2 Likes

Did you try checking your network traffic? I get similar errors, but my safe_vault.exe is still connecting when i analyze traffic through Overview - Process Hacker that was linked above.

Ive had upload speeds upwards to 1MB/s

1 Like

It seems can not make new accounts right now I guess that will come up in the logs for tomorrow; but it wont stop me from continuing to try… few more times tonight; then catch some sleep :wink:

1 Like

Thanks (and @abittner as well) for the links. Those are Windows utilities, intended to sus out the host machine they’re running on, but I’m running safe_vault on Linux.

Maybe Wireshark on my Windows box might tell the story.

1 Like

You can run wireshark on linux as well as ipperf sniffit and many others.

2 Likes

I’m to drunk the start playing with this (drunk because I had to celebrate this day). But I’ll try it out tomorrow.

Another piece of Lego, happy SAFE Network testing everyone :stuck_out_tongue:

5 Likes

up and running here, lots of ESTABLISHED connections too :smiley:

You nailed it, guys :checkered_flag: !!

9 Likes

It will likely restart tomorrow and a lot for a few days as we target churn handling more efficiently, these tests will prove really really beneficial. I hope they are quick, we get results and iterate very quickly towards … well you know what

7 Likes

hope the rest of us will get this packaged up soon :slight_smile:

4 Likes

Glad I kept trying:

just a little delayed;

7 Likes

I just got logged in too, it seems something unlocked , funny

3 Likes

got one of these on a reconnect just now;
add some more data to the logs :wink:

until the :sunrise: over the sea

OK, I have more interesting results to report:

The output of safe_vault is saying it has a routing table size of 29, after about ten minutes. That must be other nodes in my group?

That is sprinkled among many of the diagnostic “errors” that @dirvine mentions, such as deserialization failure.

Wireshark (on the Linux box, and shutting down every other network client), is capturing many “PSH” and “ACK” packets with various IP addresses.

I dare to think that it’s working.

It’s alive.

EDIT: Most of the packets are PSH and ACK, with occasional “RETRANSMISSION (malformed packet)” and KEEP ALIVE.

EDIT: By the way, I’ve found a new, probably addictive hobby: watching SAFE_VAULT packets in Wireshark.

4 Likes

Nothing can stop it.

4 Likes

Yep, I agree with you. It may look like it’s not working when you see all of those ERROR level logs but my vault is alive as well!

Don’t be discouraged by the ERROR logs. To report to the dev team what I am seeing a lot of are things like below.

failed: Only one usage of each socket address (protocol/network address/port) is normally permitted. (os error 10048)

Associated with the ip addresses hard coded into the cust config file.

1 Like

Your baby is coming and you are in the fourth stage of labor with the contractions coming fast and furious. How exciting and precious to behold! Thank you for sharing this moment with us.

5 Likes

my rust test client does work against the new 0.4.2 launcher , with the new network.

Well, it seems nothing can stop this from now… A page was turned tonight !

3 Likes